Synthesis and Characterization of Epoxidized Silkworm Pupae Oil and Its Application as Polyvinyl Chloride

More and more industries demand environmental friendliness. Silkworm pupae oil (SPO), extracted from the desilked silkworm pupae, can serve as a promising substrate alternative to use in plasticization. This study aimed to prepare epoxidized silkworm pupae oil (ESPO) and investigate their effects on the thermal stability and plasticization of polyvinyl chloride (PVC) films. A chemo-enzymatic method of ESPO was developed in the presence of Lipase SMG1-F278N and H2O2 in natural deep eutectic solvents (DESs). Lipase SMG1-F278N could initiate the epoxidation reaction effectively at room temperature with a negligible loss of activities 10 batches. A maximum oxirane value of 6.94% was obtained. The formation of oxirane ring in ESPO was confirmed by FTIR and 13C NMR spectra. Moreover, ESPO showed a better thermal stability and lower freezing point than epoxidized soybean oil (ESO). It was demonstrated that ESPO had a good frost resistance. In addition, ESPO showed a significantly improved plasticizing effect on flexible polyvinyl chloride (PVC). Compared with ESO, ESPO could increase the tensile elongation at break effectively. A significantly lower migration rate of plasticizer was observed in PVC plasticized with ESPO.


Introduction
In recent decades, epoxidation has been receiving increasing attentions in oil industries as the epoxidized fats and oils could be endowed with desirable toughness used in plasticizers [1]. It has been documented that the epoxidized soybean oil (ESO, 4.6 double bonds per triglyceride chain) enabled the significantly enhanced toughness of petro-based epoxies. Also, these epoxidized oils serve as promising eco-friendly reactive diluents [2]. The epoxidation of numerous oils has been intensively investigated and reported, including mahua oil, karanja oil, and cottonseed oil [3,4]. Today, the production of ESO has been successfully industrialized, thereby facilitating the synthesis of valuable plasticizers and polymer stabilizers [5].
Nevertheless, the industrialization of most epoxidized edible oils, including flaxseed oil, safflower oil, and rapeseed oil, is greatly limited by high costs [1]. Silkworm pupae oil (SPO), usually extracted from the desilked silkworm pupae, serves as a promising substrate alternative. Almost 650,000 tons of silkworm pupae are reportedly harvested annually in China from 2010 to 2016 [6]. Unfortunately, however, tons of silkworm pupae are disposed as industrial wastes after the extraction of nutritional silkworm pupae protein. It is well-documented that silkworm pupae are enriched sources of lipids with a content of ~ 30% [7]. Specifically, 70 ~ 80% of the fatty acid species are composed of oleic acid, linoleic acid, and α-linolenic acid (ALA) [8]. Similar unsaturation degrees (iodine value) of SPO with soybean oil are reported in previous studies [9,10]. Therefore, the application of SPO could be beneficial for the industrialization of epoxidized oils. It is demonstrated that epoxidized silkworm pupae oils (ESPO) had good thermal stability and lower coefficient of thermal expansion as epoxidized linseed oil (ELO) [11,12]. Moreover, the epoxidized fats and oils with high levels of linolenic acid exhibited better frost resistance used as plasticizers [13].
Traditional epoxidation reactions are characterized with relatively harsh reaction conditions, low catalytic efficiency, and low substrate conversion rates [14]. It is reported that the epoxidation of oils and fats using hydrolases as catalysts represents a promising solution to the above challenges [15]. Nevertheless, the enhancement of robustness of lipase is still a great concern nowadays, especially the resistance to H 2 O 2 . In order to overcome these problems, a milder alternative reaction by a chemo-enzymatic auto-epoxidation using lipase as a catalyst was proposed by Zhang et al. [16]. Constructing mutants of the enzyme was an effective way to improve the H 2 O 2 resistance [17]. Gratifyingly, it was reported that the Lipase SMG1-F278N mutant had higher catalytic activities and better reusability than the wild-type Lipase SMG1, especially for the production of propylene glycol monoester and deacidification of oils and fats with high acid values [18]. And it is well-documented that deep eutectic solvents (DESs) enabled the enhanced robustness of lipase in chemo-enzymatic epoxidation reactions [15].
In the present study, the chemo-enzymatic epoxidation of SPO catalyzed by Lipase SMG1-F278N was investigated. Furthermore, the thermal stability, cold resistance, mechanical properties, and stability of ESPO, ESO, and ELO were compared with respect to PVC plasticization.

Biological and Chemical Materials
Choline chloride and xylitol were purchased from Aladdin Chemistry Co., Ltd. (Shanghai, China). All other reagents of analytical grade were directly used without any further purification. ESO and ELO were purchased from a domestic market (RT Mart, Guangzhou, China).
Free Lipase G50 (Penicillium camemberti) was purchased from Amano Enzyme Products Company Japan (Japan). The wild-type Lipase SMG1 and Lipase SMG1-F278N were produced in our lab according to the method described previously [19]. Lipase SMG1, Lipase SMG1-F278N, and Lipase G50 were immobilized onto epoxy resins of ECR 8285 following the procedures of Li et al. [18]. The immobilized Lipase SMG1, Lipase SMG1-F278N, and Lipase G50 were obtained by filtration and then dried in vacuum at 20 °C for 12 h. The recovered enzymes were stored at 4 °C until use. The esterification activity of immobilized Lipase SMG1, Lipase SMG1-F278N, and Lipase G50, according to the Novozymes propyl laurate unit (PLU) method EB-SM1069.02, was 258, 328, and 423 U/g, respectively.

Extraction of SPO
SPO was extracted by petroleum ether using a Soxhlet extractor. After reflux at 30 °C for 8 h, the solvent was distilled off in a rotary vacuum evaporator (0.09 MPa). Then, the obtained silkworm pupae extracts were dried with anhydrous sodium sulfate for 24 h [20].

Preparation of DESs
The DESs consisting of choline chloride (ChCl, recrystallized from ethanol, filtered, and dried in vacuum) and xylitol (1:1, mol/mol) were mixed and heated at 80 °C in a rotary evaporator for 1 h. The formed colorless, homogeneous liquids were transferred into a tightly sealed bottle and then placed in a desiccator for further use.

General Procedures for Epoxidation of SPO
The procedures were according to the method described previously [16]. Firstly, SPO (4.3 g) and DESs (4.6 g) were added to a conical flask. The mixture was supplemented with 4.6 g of hydrogen peroxide (H 2 O 2 : double bonds = 2:1, mol/mol). The epoxidation reaction was initiated by addition of enzymes (20 U/g oil). The reaction medium was mixed for 12 h at 400 rpm. The products were collected at a fixed interval (2 h) and centrifuged for 3 min at 10,000 rpm. The excess H 2 O 2 was removed by ddH 2 O. The immobilized enzymes were recycled after being rinsed with excess ddH 2 O and dried in vacuum at 20 °C.

A Scale-up Reaction
A scale-up reaction (~ 100-fold) was carried out under the optimized conditions. 108 g of SPO, 115 g of DESs, and 115 g of hydrogen peroxide (H 2 O 2 : double bonds = 2:1, mol/mol) were added to the reaction vessel. After the addition of Lipase SMG1-F278N (20 U/g oil), the epoxidation reaction was catalyzed at 30 °C for 8 h.

Analysis of Fatty Acid Composition by Gas Chromatography
The fatty acid composition was determined as fatty acid methyl esters (FAMEs) according to the standard method described in ISO 5509:2000 (E). The FAMEs were analyzed by gas chromatography-flame ionization detector (GC-FID) equipped with a CP-Sil 88 capillary column (60 m × 0.25 mm, 0.2 µm film thickness; Dikma Technologies, Beijing, China) [20].

Characteristics of ESPO
The formation of oxirane ring in ESPO was confirmed by Fourier transform infrared spectrometer (FTIR) and carbon-13 nuclear magnetic resonance ( 13 C NMR) spectra. The FTIR spectra of the products were recorded on an FTIR (Thermo Scientific, model Nicolet 6700, Waltham, MA, USA) with a resolution of 4 cm −1 . Data were acquired using the FTIR software (Thermo Scientific, OMNIC series suite, Waltham). The 1 H and 13 C NMR spectra of SPO and the derivatives (10 ~ 15 mg/mL in CDCl 3 ) were recorded by a 600 MHz Brucker NMR spectrophotometer.
The thermal degradation of the obtained polymers was determined by thermogravimetric analysis (TGA) using Mettler-Toledo TGA/SDTA 851e. TG analysis was performed using a thermal weightlessness analyzer (TG209-F3, NETZSCH Co., Germany) in the temperature range of 25-700 °C with a rate of 10 °C/min under N 2 (20 mL/min).
The curing behaviors of epoxidized oils were characterized by a differential scanning calorimeter (DSC) (Netzsch Geratebau, model 204 C, Germany), in the temperature range of − 60 to 40 °C using the liquid nitrogen.

The Mechanical Properties of PVC Samples
In the study, the tensile strength and elongation at break of PVC films made of ESPO were investigated to evaluate the comprehensive performance of various epoxy greases according to the American Society for Testing and Materials (ASTM) standards (ASTM D638).

The Thermal Stability and Migration Tests of the PVC Samples
The plastisol was prepared by mixing 50, 60, 70, 80, 90, and 100 phr of plasticizer with 100 phr PVC for 15 min, respectively. Subsequently, the mixture was compounded by double-roller blending rolls (SK-160B, Shanghai Rubber Machinery Factory) at 165 °C for 8 min. The sample was pressed on a flat plate vulcanizing machine (HY-100TA, Shanghai Hengyu Instrument Co., Ltd.) at 170 °C for 6 min to make film with a thickness of 5 mm. The final film was obtained after the temperature of the flat plate vulcanizing machine dropped to 50 °C.
The thermal stability of PVC samples was tested according to the Congo red method in GB/T 2917. . PVC samples were cut into 2 mm × 2 mm and placed in test tubes. Then, the test tubes were incubated in an oil bath at 170 °C. The time required for color changes of the Congo red test paper represented the thermal stability.
The migration tests were performed according to the following procedures [12]. Firstly, all samples were immersed in n-hexane at 50 °C for 2 h. The weight loss was measured gravimetrically. It is considered that the weight of residue extracts present in n-hexane is equal to the plasticizer extracted.
Secondly, the mass loss of a plastic piece placed between two absorbent films was measured according to ISO 177. Samples were placed on two plates (3 mm thick) of low density polyethylene (LDPE) at 70 °C for 24 h with a circular cylinder (1 mm thick) of PVC plasticized with ESPO in between. The amount of plasticizer migrated was determined according to the weight differences before and after tests.
The third migration test was conducted using activated carbon as per ASTM D 1203-94 test method B. A wire cage was used to prevent from direct contact between the plasticized PVC and carbon. The migrated volatile components were measured after the circular cylinder was placed at 100 °C for 24 h [12].

Statistical Analysis
All experiments were performed in triplicate. The data were presented as mean values ± standard deviations. One-way analysis of variance was performed using the SPSS software (version 14.0 demo, SPSS Inc., Chicago, IL, USA).

Fatty Acid Composition of SPO
The fatty acid composition of SPO was presented in Table 1. ALA was almost 36% of the total fatty acid composition. Studies have also found that ALA was the most abundant fatty acid in SPO, accounting for about 34 ~ 38% [8,20]. The total concentration of unsaturated fatty acids was 75.13%, and the iodine value of SPO was 138.7. It was observed that the iodine value of SPO reported herein was consistent with that of soybean oil [21]. Moreover, it was demonstrated that ALA was the primary unsaturated fatty acid. Therefore, SPO

Synthesis of ESPO
The superiority of partial glyceride lipase as a catalyst for triglyceride epoxidation has been proved in previous studies [15]. In this study, a mutant of Lipase SMG1 (SMG1-F278N) was selected to synthetic ESPO. Compared with commercial Lipase G50 and wild-type Lipase SMG1, Lipase SMG1-F278N exhibited higher SPO epoxidation activities (Fig. 1). The epoxidation reaction reached to the equilibrium in less time using Lipase SMG1-F278N, with a higher final oxirane value (> 6.9) being observed [22]. Therefore, Lipase SMG1-F278N has superior catalytic activities for epoxidation of SPO.
The previous studies demonstrated that the immobilized SMG1-F278N had a higher temperature tolerance, with the maximal catalytic activities at 30-35 °C [18]. In our study, an increasing reaction rate was observed from 20 to 30 °C (Fig. 1A). And the low concentration (0-0.5 mol/L) of peroxyacid would not result in a significant loss of enzyme activity. The thermostability of enzymes was primarily contributed by hydrogen bonds, ion pairs, and hydrophobic bonds [17]. Overall, it could be concluded that high temperatures (> 35 °C) and peroxyacid produced both might destroy these structures leading to the loss of catalytic activities [14]. Fig. 1 Effects of temperature on epoxidation activities of different enzymes. ( *) indicated that the oxirane value at different temperatures was significantly different from 0 h; (* *) indicated that the oxirane value at different temperatures was significantly different from 2 h; (* * *) indicated that the oxirane value at different temperatures was significantly different from 4 h; (* * * *) indicated that the oxirane value at different temperatures was significantly different from 6 h; (* * * * *) indicated that the oxirane value at different temperatures was significantly different from 8 h; (* * * * * *) indicated that the oxirane value at different temperatures was significantly different from 10 h

Scale-up Reaction and Reusability of Immobilized SMG1-F278N
It is generally acknowledged that immobilized lipases could be recycled easily and reused compared with free counterparts. However, the recovery of partial glyceride lipases used for triglyceride epoxidation has not been investigated. In the present study, we explored the reusability of immobilized SMG1-F278N at low temperatures (25 °C and 30 °C) and the scale-up reactions (Fig. 2).
After being reused for 10 consecutive batches at 25 °C and 30 °C, the immobilized SMG1-F278N still had 94% and 96% of the original catalytic activities, with oxirane values 6.49 and 6.67 being observed. It was documented that the regenerated Novozyme 435 (lipase from Candida antarctica) dramatically loses its epoxidation catalytic activity if the reaction temperature exceeds 25 °C [23]. The small amount of peroxyacid (< 0.24%, w/w%) produced in the reaction generally could not lead to enzyme inactivation. The exogenous acids (acetic acid or other acids) added largely inhibited the hydrolysis of oils and fats. Meanwhile, these acids served as effective substrates for production of peroxyacid. However, these exogenous acids could not only lead to decreased stabilities of lipase, but increased the complexity of product separation and purification [24]. This study found that Lipase SMG1-F278N could catalyze the epoxidation of unsaturated fatty acids of SPO effectively without exogenous acids. In addition, the epoxidation reaction reported herein, to a large extent, avoided the hydrolysis of glycerides and improved the enzyme stability. The enzyme could be reused for at least 8 times without significant loss of activities.
The changes of oxirane values of ESPO products are illustrated in Fig. 2B. It was found that the scale-up (by 25 times) reaction reached the equilibrium after 6 h with an oxirane value of 6.94%. Overall, this epoxidation reaction could be scaled up with a good reproducibility.

Characterization of ESPO
The physicochemical indices of ESPO were determined (Table 2). Obviously, the acid value, iodine value, and heating loss of ESPO were lower than that of ESO and ELO, while the content of epoxy groups of ESPO was higher than ESO. Similar densities were observed between ESO and ELO. These physicochemical parameters indicated that ESPO could serve as a promising alternative of ESO as plastic stabilizers and plasticizers [12,25]. The acid value of ESPO increased from 0.14 mg/g KOH to 0.47 mg/g Fig. 2 Reaction batch and reaction scale on the epoxidation reaction. The letter "a" indicated that there was no significant difference in the oxirane value of the reaction batch at 25 °C or 30 °C (A); the different letters indicated that there was significant difference in the oxirane value of the different time (B) KOH after epoxidation. This increment could be explained by the increased levels of free fatty acids produced via hydrolysis of diglycerides. The acid value of ESPO was less than 0.5 mg/g KOH, indicating that the triglycerides were not hydrolyzed.

FTIR and NMR Analysis
The FTIR spectra of SPO and ESPO are shown in Fig. 3. The absorption band at 3009 cm −1 was characteristic of = C-H (Fig. 3A). The ESPO exhibited an absorption peak at 837 cm −1 (Fig. 3A), which was attributed to the epoxy group. Moreover, the decrease of peak intensity at 3009 cm −1 indicated the consumption of double bonds [24], while the peak intensity at 837 cm −1 increased dramatically with increased epoxidation degrees. Therefore, SPO was successfully epoxidized by Lipase SMG1-F278N. Figure 4 shows the 13 C NMR spectrum of the SPO and ESPO. The peaks at δ 172.5-173.5 ppm of SPO and ESPO were attributed to the ester carbonyls in triglycerides. The characteristic peak of CH = CH was observed at δ 127.1 ppm. Similarly, Li also determined the formation of epoxidation products according to the characteristic signal of C = C [18]. Obviously, a new peak at δ 31.5 ppm corresponding to the epoxy group was found in the epoxide product. Therefore, it was concluded that the epoxy groups were obtained by Lipase SMG1-F278N effectively.

Thermal Characterization
It is generally acknowledged that the thermo resistance of epoxy grease per se could contribute to the thermal stability of PVC and other plastic products. The thermal degradation kinetics of SPO, ESO, ELO, and ESPO are shown in Fig. 5. As can be seen from Fig. 5A, the thermal decomposition temperature of SPO increased from 374 to 416.3 °C after epoxidation, which is consistent with the epoxidized vegetable oil reported previously [26]. Similar thermal decomposition temperatures of ESPO (416.3 °C) and ELO (411.5 °C) could be observed from Fig. 5B, which were higher than that of ESO (402.3 °C). Therefore, ESPO could improve the thermal stability of plastic products significantly. It has been demonstrated that the frost resistance of epoxy grease is closely related to solidification temperatures [13]. In this study, DSC was used to investigate the solidification temperature of ESPO. It was found that the initial solidification temperature of ESPO was − 3 °C, which was close to ELO (− 5 °C) and lower than ESO (9 °C) (Fig. 6). Therefore, ESPO has better frost resistance than ESO as a plasticizer.

Effect of Plasticizer Addition on PVC Mechanical Properties
Some vital processing parameters including curing formulations and mechanical characterization are usually assessed to evaluate the flexibility of PVC.
As shown in Fig. 7A, the tensile elongation of ESPO on PVC was higher than that of ESO with the plasticizer content of > 60 phr. The tensile strength of PVC was significantly reduced with the plasticizer content of > 80 phr (Fig. 7B). Notably, similar mechanical properties could be found with respect to the PVC sheets plasticized with ESPO and ELO [11]. This observation could be explained by that the majorities of epoxy glycerides in ESPO and ELO were epoxy linolenic acid glycerides [27]. Therefore, ESPO has better plasticizing effects on PVC than ESO.

Thermal Stability and Migration Test of PVC Samples
The results about thermal stability and migration test of PVC samples are shown in Table 3. The results demonstrated that PVCs (80 phr) plasticized with ESPO and ELO showed close thermal stability times of 30 min and 29 min, respectively, which were higher than 26 min of ESO. Therefore, ESPO contributed to better stabilities of PVC than ESO. The good thermal stability of epoxy oil was attributed to its long carbon chain, which made its dispersion in PVC matrix faster and more uniform [28]. The migration tests demonstrated that the migrate rate of ESPO (− 0.83 ± 0.02%) from PVC was significantly lower than ESO (− 1.02 ± 0.05%). Fenollar thought uniform structures with the plasticizer absorbed by PVC particles have the minimum values of migration of plasticizer [12]. Therefore, ESPO serves as a promising alternative of ESO as the raw material for production of plasticizers.

Conclusions
In the present study, an effective chemo-enzymatic route was proposed using Lipase SMG1-F278N as catalysts for ESPO in DESs. The preliminary results showed that a high catalytic efficiency for production of ESPO was observed due to the enhanced robustness of Lipase SMG1-F278N in DESs. Compared with ESO, the ESPO reported in this contribution exhibited better thermal stability and frost resistance. In addition, a relatively low migration rate of plasticizer was observed in PVC plasticized with ESPO. Overall, ESPO is expected to serve as a promising alternative to ESO used for packaging films.